Annu. Rev. Astron. Astrophys. 1984. 22: 445-70 Copyright © 1984 by Annual Reviews. All rights reserved

4.5 Discussion

Because graphic evidence that suggests ongoing interaction between galaxies and their environment has been collected for a variety of individual objects (37, 72, 72a, 76, 99), there has been a tendency toward interpreting the H I deficiency as the result of presently active mechanisms, rather than the remnant effect of earlier, now exhausted periods in the evolution of clusters. There has not been, however, a general consensus. As documented by Dressler (32), it is difficult to account for all the aspects of the morphological type dependence on local density purely in terms of secular stripping mechanisms. Hence, conditions at some early phase must be responsible for most of the differentiations: while bulges collapse rapidly, disks may do so slowly (50) and continue to accrete gas from external reservoirs; the tidal stripping of those reservoirs, at early stages of cluster formation, would arrest the growth of disks (74), thus explaining the preponderance of lenticular systems. Could the ``arrested growth'' hypothesis account for the observed H I deficiencies? This view has been favored by some (11, 70).

One obvious direction to search for corroborating evidence of the deficiency measurements is in the colors of the H I-poor cluster galaxies. In fact, colors of Virgo spirals are found to be redder than those of spirals in the field (70, 104), and differences have been reported also for other clusters (14, 43). However, in Virgo, H I-deficient spirals exhibit a large difference in their (U - V) colors with respect to nondeficient spirals (41, 44), a result that suggests a relatively sudden and recent arrest of star formation in the first group and that is at odds with the arrested growth hypothesis. Disk H + [N II] line emission also reveals a striking difference between Virgo and field galaxies (70); on the other hand, nuclear H emission does not appear to be significantly depressed in Virgo spirals (104), suggesting that the cluster environment has preferentially affected the outer regions of the galaxies. This result is in qualitative agreement with the observed reduction of H I disks discussed in Section 4.3. The morphological appearance of the spiral pattern has been invoked to define a class of anemic spirals (109), suggested to represent a transitional phase of a spiral galaxy when star-forming activity has ceased. Although the morphological classification involves a measure of subjectiveness, anemic spirals, and the more extreme cases of ``smooth-arm'' spirals (117), have been shown to be very H I-deficient objects.

The pattern of H I deficiency is not equally as pronounced in all clusters. some authors (15) have reviewed the results, concluding that evidence for a global sweeping mechanism at work is present only in Coma. Others have found that the magnitude of the effect depends on cluster properties, most notably on the cluster's X-ray luminosity (46) (as can be reconstructed from Table 1). Strong X-ray emitters, like Coma and A2147, present the highest fraction of deficient spiral galaxies. The comparison between A2147 and A2151 is of particular interest: while the space density of spirals is higher in A2151, and the velocity dispersions are comparable, a circumstance that would favor a higher collision rate and therefore more effective collisional stripping in A2151, the H I deficiency is more pronounced in A2147, which is dominated by a cD galaxy and a healthy ICM.

Another clue could be provided by comparisons of H I deficiency with a galaxy's velocity relative to the ICM and with the cluster's X-ray temperature, as they may permit distinctions between the efficiency of ram pressure and evaporative processes. In analyzing the velocities of individual galaxies, the lack of knowledge of a galaxy's exact location and velocity vector introduces a great deal of uncertainty. A correlation supportive of ram pressure stripping at work has been obtained for the Virgo cluster, although no confirmation in other clusters has been found (44, 106). Data on X-ray temperatures of the ICM are available for only five of the clusters well studied in H I, and the paucity of the statistics allows little speculation; it is interesting to note, however, that A2147 has the lowest of cluster X-ray temperatures, yet its H I deficiency is quite pronounced.

The various pieces of circumstantial evidence tend to support a picture of ICM-induced gas removal from galaxies in clusters. The large fraction of H I-poor galaxies in several clusters suggests that the efficiency of the gas removal processes is higher now than it was a few billion years ago (46), a conclusion also reached by Gisler (48) on theoretical grounds. Schemes based on evolutionary models that may distinguish a lenticular galaxy that originated from a stripped spiral from one whose morphology originated in earlier evolutionary stages have not yet been developed. However, there appears to be an enhanced fraction of S0s, and a matching depression of the fraction of spirals, at any given value of the local density of galaxies in clusters that are strong X-ray emitters (32). Furthermore, de Freitas et al. (28a) have found that the axial ratios of S0 galaxies tend to be higher in regions of high galaxian density than in the field; they identify the flat cluster S0s as swept spirals. These statistical results suggest that morphological evolution, secularly imposed by environmental factors, is reinforcing the characteristics of the segregation that may have been established ab initio (46).